🧬 Cloning Skill
A Cloning Skill is a protocol that takes a validated template and rebrand for a new client or niche: swap the logo, copy, brands, prices, reviews, and accent color—without rebuilding the structure from scratch.
✓ When to use it
- ✓You already have a template that works
- ✓Needs to serve several similar clients
- ✓You want to keep the structure and just change the branding
✗ When to avoid it
- ✗The new project has a different structure
- ✗You haven’t validated the original template yet
- ✗The content changes more than the brand
cloning / rebrand protocol
Use a Cloning Skill neste template. Re-marque para o cliente abaixo, trocando: logo, copy, nome de marca, preços, avaliações e cor de acento. Mantenha a estrutura intacta. Cliente: [nome] · nicho: [nicho] · cor: [hex]
🔄 Transformer Skill
A Transformer Skill converts a site built with a framework into a single self-contained HTML file — everything is built in (inline CSS and JS) — ready to paste into platforms such as GoHighLevel or WordPress.
📦 Deliver where the client already is
Many clients already have their site on GoHighLevel or WordPress. Instead of migrating everything, you turn your page into a single block and paste it there. One file, zero external dependencies.
framework → self-contained HTML
Use a Transformer Skill: converta este projeto num único HTML self-contained (CSS e JS inline, sem dependências externas), pronto para colar dentro do GoHighLevel/WordPress.
🐞 Ultimate Debugger Prompt
O Ultimate Debugger Prompt is a structured template (in XML) to debug errors with AI in a disciplined way: the model generates predictions, investigates, finds the root cause, and gives step-by-step instructions. It's a third-party open-source tool.
Predictions
AI lists hypotheses about what might be causing the error before investigating.
Investigation → root cause
Checks each hypothesis and pinpoints the real source of the problem.
Step-by-step instructions
Provides the detailed fix in order, ready to apply.
XML debug template · open-source
<debug> <erro>[cole a mensagem de erro]</erro> <contexto>[o que você estava fazendo]</contexto> <tarefa>Liste predições, investigue, ache a causa-raiz e me dê instruções passo a passo para corrigir.</tarefa> </debug>
🔌 MCP Config
O MCP Config is a ready-made configuration for MCP servers— n8n, Notion, Pinecone, Supabase, Vercel, and Fireflies — with key placeholders. It's what connects your agent to external tools.
| MCP | What for |
|---|---|
| n8n | trigger and read automations |
| Notion | read/write pages and databases |
| Pinecone | vector search / memory |
| Supabase | database and auth |
| Vercel | deployment and domains |
| Fireflies | meeting transcripts |
🔑 Be careful with keys
The file comes with placeholders (e.g., SUA_CHAVE_AQUI). Replace these with your values, and never push real keys to public repositories.
6 MCPs with key placeholders
Conecte o MCP do [Supabase] usando esta config. Me oriente passo a passo onde colar a configuração e onde substituir a chave (placeholder). Depois, confirme que a conexão está funcionando.
📋 claude.md / Gemini.md (CODA)
The template for claude.md / Gemini.md is the starting point for the “C” in CODA’s Configure: the context file with tech stack, directories, commands, success criteria + how to test. Start every project from it.
# [Nome do projeto] ## Stack - ... ## Diretórios-chave - ... ## Comandos - rodar: ... - testar: ... ## Critério de sucesso - ... ## Como testar 1. ...
CODA context file
Fill out this claude.md template for my project. Ask questions about the tech stack, directories, commands, and—most importantly—success criteria and how to test, until the file is complete.
✅ QA Pre-Flight Checklist
O QA Pre-Flight Checklist is a QA and stress-testing protocol by phase, which classifies each finding into three severity levels— Critical, Notice e OK. It's your check before "taking off" (publishing).
blocks use · cannot be published
works, but improves later
passed the test
QA by phase · Critical/Warning/OK
Run the QA Pre-Flight Checklist on this project. Perform the stress test in phases and classify each finding as Critical, Warning, or OK. List the Critical findings first and say whether I can publish.
⚡ Prompt library
Finally, the Prompt library: a collection for any stage of CODA. problem→SOP, Plan Mode, research→research.md, design inspiration, review/verify, screenshot-based troubleshooting, connect MCP, env var and the "less is more" reminder.
becomes a procedure
plan ahead
saved research
references
check everything
image-based debugging
connect tools
secure variables
short prompt
collection for all of CODA
Transforme o problema abaixo num SOP (procedimento operacional padrão): passos numerados, claros e repetíveis, com critério de "feito" em cada um. Problema: [descreva]
🔬 Practical example
Let’s get the library running: download the Cloning Skill and use it to switch the niche of a template you already have.
- Download cloning-protocol.md in topic 1.
- Point the skill to a validated template of your own.
- Ask it to reschedule for a new niche (e.g., from “clinic” to “gym”): logo, copy, prices, reviews, and color.
- Check that the structure stayed intact and only the branding changed.
In minutes, you have a second version ready — that’s the power of reusing artifacts from the library.
🏋️ Exercise
Choose 1 skill from this library and adapt it to your project:
- Download the artifact (skill, config, or prompt).
- Edit it for your context — replace placeholders, niches, commands, or criteria.
- Save the your version with a proper name and a test in the project.
Tip: start with claude.md (topic 5) — it serves as the foundation for all the other CODA steps.
⚡ Ready-made prompts
Copy, paste into your AI assistant, and adapt to your use case.
Adapt the skill below to my project: replace the placeholders, adjust the niche/context, and send me back my version, ready to save. Skill: [paste the content] · My context: [describe it]
Anexei um screenshot do problema. Descreva o que está errado na tela, liste as causas prováveis e me dê o passo a passo para corrigir.
Configure a chave [SERVIÇO] como variável de ambiente (env var) neste projeto, sem deixá-la hardcoded no código. Me mostre onde colocá-la e como referenciá-la com segurança.
✅ Module summary
End of the learning path:
You’ve completed Mastery. Return to the beginning to review any track whenever you like.